Ecosystem Services 6 (2013) 72–81

Contents lists available at ScienceDirect

Ecosystem Services

2212-04http://d

n CorrSustainaNetherl

E-m

journal homepage: www.elsevier.com/locate/ecoser

Payments for ecosystem services and rural development: Landowners'preferences and potential participation in western Mexico

Arturo Balderas Torres a,b,n, Douglas C. MacMillan c, Margaret Skutsch a,d, Jon C. Lovett a,e

a CSTM, Twente Centre for Studies in Technology and Sustainable Development, University of Twente, Postbus 217, 7500 AE Enschede, Netherlandsb Ingenieria Ambiental, PTI Instituto Tecnológico y de Estudios Superiores de Occidente (ITESO), Tlaquepaque, CP 45090, Méxicoc Durrell Institute of Conservation and Ecology, School of Anthropology and Conservation, University of Kent, Canterbury, CT2 7UH, England, UKd Centro de Investigaciones en Geografía Ambiental, Universidad Nacional Autónoma de México (UNAM), Antigua Carretera a Pátzcuaro 8701, CP 58190Morelia, Michoacán, Méxicoe School of Geography, University of Leeds, Leeds, LS2 9JT, UK

a r t i c l e i n f o

Article history:Received 10 August 2012Received in revised form6 March 2013Accepted 28 March 2013Available online 8 May 2013

Keywords:Livelihoods approachMarket-based mechanismsIncentive programsREDD+La Primavera Biosphere Reserve

16/$ - see front matter & 2013 Elsevier B.V. Ax.doi.org/10.1016/j.ecoser.2013.03.002

esponding author at: CSTM, Twente Centre fble Development, University of Twente, Posands. Tel.: +31 53 489 3203.ail address: a.balderastorres@utwente.nl (A. B

a b s t r a c t

Incentive-based mechanisms can contribute to rural development and deliver environmental services,but need to be attractive to landowners and communities to ensure their participation. Here we studythe views of landowners and agrarian communities (ejidos) from central Jalisco in Mexico to identifycharacteristics that payment for environmental services (PES) programs conserving/enhancing forestcover could include in their design. A choice experiment was applied to 161 landowners and ejido-landowners. Results show that importance and dependency on cash payments can decrease ifinterventions to promote local development through improved health and education services andgeneration of employment and productive projects are included. Responses indicate that communalforested areas in ejidos would be most likely to enroll into PES. In some cases grasslands could beafforested. Agroforestry practices providing other environmental services could also be implemented(e.g. windbreaks). Potential enrollment is lower in agricultural and peri-urban areas due to higheropportunity costs. Higher payments favor enrollment but may compromise the program's efficiency sinceaggregated cash-flow over long periods can exceed the present value of the land itself in some areas.Offering a mix of cash and non-cash benefits based on local developmental needs might be the best wayto promote participation in PES.

& 2013 Elsevier B.V. All rights reserved.

1. Introduction

In recent years interest has increased in improving our under-standing of the institutional and economic arrangements neededto maintain ecosystem services (Balvanera and Cotler, 2007). Theseecosystem or environmental services include regulation, habitat,production and information functions (de Groot, 1992; de Grootet al., 2002). They are present at local or transboundary scales(López-Hoffman et al., 2010), and include benefits such as protec-tion of water catchments and biodiversity habitat, sites forrecreation and climate change mitigation (Canadell and Raupach,2008). However, many of these benefits are public goods, hencelandowners and communities protecting the ecosystems providingthem are often not compensated. In other words, there is a marketfailure as the value of the services is not recognized by theconsumers and the producers may opt for land uses other than

ll rights reserved.

or Studies in Technology andtbus 217, 7500 AE Enschede,

alderas Torres).

those that provide the services. As a result, environmental servicesand functions can be lost when they are not factored into land-owners' land use decisions (e.g. Grieg-Gran et al., 2005; Pagiolaet al., 2002).

Recent environmental policies have aimed to solve the problemof market failure by creating positive incentives for landownersthrough a number of approaches including market based mechan-isms and subsidies (e.g. Kinzig et al., 2011; Landell-Mills andPorras., 2002; Muradian et al., 2010). These policies seek tosupport practices that maintain and enhance provision of envir-onmental services through payments to the service providers.Usually these payments are designed to match the opportunitycosts of some alternative land uses that imply reductions in theflow of environmental services (Pagiola and Platais, 2007). Pro-grams of payments for environmental services (PES), carbonsequestration projects financed through carbon markets andpolicies to incentivize reductions in emissions from deforestationand forest degradation in developing countries (REDD+) areexamples of these schemes. It is important to identify the contextwithin which these policies are to be applied. Lack of propertyrights over land, incomplete information on management

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–81 73

practices, benefits and costs, or lack of access to capital markets(e.g. credit) need to be assessed first to evaluate the appropriate-ness of a PES or incentive-based approach in a policy mix (Engelet al., 2008).

Incentive-based schemes are often considered to have thepotential to contribute to poverty alleviation and poverty preven-tion (Daw et al., 2011; Grieg-Gran et al., 2005; Landell-Mills andPorras, 2002; Muradian et al., 2010; Pagiola et al., 2002; Wunder,2005). Contributions to poverty alleviation come from directbenefits derived from maintenance and access to the environ-mental services; and co-benefits associated with the projects suchas cash-income and labor opportunities (Daw et al., 2011). Thesetypes of schemes may only succeed in the long term if the benefitsgenerated are considered adequate by local communities andlandowners (Engel et al., 2008; Grieg-Gran et al., 2005; Landell-Mills and Porras, 2002). Theoretical and empirical research hasshown that rural landowner participation in incentive programs isinfluenced by different factors including the characteristics of theschemes, the characteristics and attitudes of the individuals andthe activities developed on the land (e.g. Ma et al., 2012; Puttenet al., 2011).

In addition to direct environmental services and direct pay-ments, other benefits can also be considered in a broader develop-mental context in the valuation of ecosystem services. According tothe livelihoods approach, the level and potential for development isrelated to the levels and changes in social, human, productive,financial and natural capital (e.g. Carney, 1998). As described byLandell-Mills and Porras (2002) PES can have positive effects in thefive dimensions of capital through enhancement of local environ-mental services which are enjoyed locally by the providers; bystrengthening local institutions; building capacity and transferringspecific know-how; by the creation of specific infrastructure asso-ciated with human and productive capital (e.g. tree nurseries, roads,health care centers, schools); and through cash payments andincreased access to markets that increase financial capital. Yetmost incentive programs offered as part of PES have focusedprimarily on financial compensation only, for example the MexicanPES program (Muñoz Piña et al., 2008). In a review of 47 PESschemes worldwide for watershed services, Brouwer et al. (2011)report that in around 70% of the cases analyzed the main mode ofcompensation was through cash payments. Low payments mayrepresent small contributions to the landowner's total income; andadoption of PES is likely to indicate the low opportunity costs ofenrolling in the programs (e.g. PES contributed from 1% to 3% ofyearly income to participants in Coatepec Mexico) (Scullion et al.,2011).

It is possible to design programs to conserve environmentalservices based on the livelihoods approach. An example of such aprogram is Bolsa Floresta in the Brazilian Amazon (Viana, 2008).This program provides four types of support to participantsengaging in forest conservation: a fixed monthly payment perhousehold; community payments for income generation activities;community payments for the development of local infrastructuresuch as water sanitation, health or education centers or roads; andresources to strengthen local institutions and organizations(Newton et al., 2012; Viana, 2008). It is a governmentally mixedprogram in which international development agencies and locallybased non-governmental organizations also participate (Newtonet al., 2012).

The aim of the study presented here is to explore the potentialparticipation of landowners in a project using incentive-basedmechanisms for the conservation of forests with a region inwesternMexico as a case study. We developed a choice experiment toelucidate the effect that cash payments, project duration andpotential access to other co-benefits promoting local developmentas part of the project, might have in potential participation. Based

on participant responses we discuss the implications for PES designand the contribution to rural development and provision ofenvironmental services. First we present background informationabout participation of landowners in incentive-based programs.Thenwe describe the methods used and study area, followed by theresults and discussion. Finally, we present our conclusions.

2. Background

2.1. Forestry incentive-based programs and landowner participation

Concern over relative scarcity of natural services is leading topromotion of policies for the creation of incentives to conservediminishing ecosystem functions (Kinzig et al., 2011; Kroeger andCasey, 2007). PES was initially defined by Wunder (2005) as avoluntary agreement between at least one buyer and one seller,based on provision of a specific environmental service. A morerecent definition considers PES as the transference of resourcesamong social actors, through markets, incentives or subsidies, toalign the management of natural resources with a wider socialinterest (Muradian et al., 2010). Benefits for the providers partici-pating in PES can include cash and in-kind payments, capacitybuilding, income diversification, reliable and stable payments andimproved organization (Grieg-Gran et al., 2005). As opposed tocommand and control policies, PES and market based mechanismsgenerally rely on voluntary participation of landowners in envir-onmental management, thus the incentives offered by theseprojects need to be of a type and level which attracts participation(Engel et al., 2008). Economic theory assumes that if landownersdecide to participate in incentive-based programs, it is because thebenefits received make them at least as well-off as they werewithout the program (Grieg-Gran et al., 2005).

Previous research indicates that there is a range of factorspromoting or preventing participation of landowners in incentive-based schemes. Theoretical and empirical studies indicate a greaterpotential for participation when payments are higher (e.g. Dickinsonet al., 2012; Lynch et al., 2002; Markowski-Lindsay et al., 2011; Puttenet al., 2011). However there may be landowners already interested inconservation who are willing to participate and for whom theflexibility of the scheme might be more important (e.g. Horne,2004; Putten et al., 2011; Scullion et al., 2011; Stevens et al., 2002).In general, landowners prefer simpler projects without additionalityrequirements (Markowski-Lindsay et al., 2011), with lower compli-ance costs (Putten et al., 2011), not requiring specific managementprograms and not including penalties for non-compliance or earlywithdrawal (Dickinson et al., 2012; Markowski-Lindsay et al., 2011).Schemes offering upfront payments might be preferred (Putten et al.,2011) as they maximize net present value (Balderas Torres et al.,2010). There is also a preference for shorter project lengths since thisallows more management flexibility in the future, especially wheninter-generational aspects are considered (Dickinson et al., 2012;Horne, 2004; Markowski-Lindsay et al., 2011; Putten et al., 2011;Stevens et al., 2002).

Participation in PES schemes has been linked with higher income,off-land income, youth and education level and low levels of house-hold debt (Dickinson et al., 2012; Dupraz et al., 2003; Greiner et al.,2003; Lambert et al., 2007; Loftus and Kraft, 2003, Petrzelka et al.,2012; Putten et al., 2011). Positive participation has also been relatedto larger property size (Kilgore et al., 2008; Ma et al., 2012, Puttenet al., 2011), and also previous knowledge of the program andpositive attitudes towards the environment (Ma et al., 2012; Puttenet al., 2011). Landowners may have different reasons for owning theirland and they may be positive about pro-conservation incentiveprograms as long as these do not conflict with their core interests orbusiness objectives (Church and Ravenscroft, 2008; MacMillan and

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–8174

Phillip, 2010). In areas with more pro-development land use regula-tions, relatively lower participation in incentive-based conservationprograms is expected since more profitable activities would beconsidered by landowners (Markowski-Lindsay et al., 2011).

2.2. PES in Mexico and land tenure

The federal Mexican government started the national PESprogram in 2003 to compensate landowners and communitiesmonetarily for conserving forests (Muñoz Piña et al., 2008; RicoGarcia-Amado et al., 2011). The payment level was estimated bythe government according to a deforestation risk map and doesnot cover opportunity costs (Corbera et al., 2007; Rico Garcia-Amado et al., 2011). Incentives to landowners consist of yearlypayments over five years; participation can be renewed dependingon performance and the budget available. In addition to the federalprogram, other PES programs have been created including anumber developed jointly with the private sector and others atthe municipal and state levels. The payment levels range from$22.2 per hectare per year in the federal program to $111 for thePROBOSQUE program in the State of Mexico (Estado de México,2011; Muñoz Piña et al., 2008; all monetary figures presented inthis paper are in U.S. dollars, the exchange rate used is $13.50Mexican pesos per dollar).

In Mexico land tenure comprises public (government property),private and community or ejido property regimes. Nationally,approximately 53% of the territory is ejido land, 37% is privateproperty, 4% public property and the remaining corresponds toother regimes (e.g. agrarian colonies) (Reyes-González et al., 2012).In the private property regime potential enrollment into PES orother conservation programs constitutes an individual decision.Under the ejido or community regimes, a group of individuals isgiven entitlement to an area of land by presidential decree. Onepart of this is divided into individual parcels, usually withpotential for agricultural activities, so that each ejidatario decideshow to use this for him/herself. Another part of the land becomes

Fig. 1. Study area showing La Primavera and the Metropolitan Area of Guadalajara. L

a common area and usually this remains as forest. Decisionsregarding the activities in common areas, such as enrollment intoPES, are taken communally in the official ejido assemblies. As arule, ejidatarios transfer the entitlement of their individual plots tosingle heirs only after they die (the parcels are therefore notsubdivided). However, individual certificates can also be trans-ferred and sold to other people provided this is approved by theejido general assembly. Since 1992 when the Constitution waschanged, these parcels can be unincorporated from the ejido(privatized) (Cornelius and Myhre, 1998; Rico Garcia-Amadoet al., 2011). This reform permitted more rapid land use changein specific areas, such as those close to the coast with potential fortourism, or for housing where the ejidos are adjacent to urbanareas. Although the figures are dynamic due to reforestation,deforestation and privatization of land, previous studies indicatethat ejidos own around half the territory of the country and 80% ofMexican forests (Barnes, 2009; Rico Garcia-Amado et al., 2011).There are parts of the territory remaining as government property.In a few cases public areas result from the public purchase orprivate donation of land for the creation of natural protected areas.In these cases, the provision of environmental services can makeuse of policy options other than incentive-based mechanisms(e.g. management of protected areas).

2.3. Case study

The study area is the La Primavera forest in Jalisco, togetherwith surrounding areas of forest patches up to 80 km away inorder to include the wildlife corridors connecting La Primavera toother forested areas (Fig. 1). The total area is around 1.5 millionhectares, of these, around 728,800 ha are forested, 155,480 ha aregrasslands and 560,700 ha are under agriculture (INEGI, 2010a).The main vegetation types in the area are oak-pine mixes, drytropical deciduous forests and grasslands interspersed with agri-culture. La Primavera forest (30,500 ha) was established in 1981 asa protected area and in 2006 it was declared a Biosphere Reserve

and use and vegetation types grouped into broad classes based on INEGI, 2010b.

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–81 75

by MAB-UNESCO (CONANP, 2000; UNESCO, 2011). On the easternborder is the city of Guadalajara, the capital of Jalisco and thesecond most populated metropolitan area in Mexico, with around4.4 million habitants (INEGI, 2010b). The metropolitan area ofGuadalajara includes eight municipalities (INEGI, 2005). Proximityto and accessibility from the city increases the perception andvaluation of environmental services (e.g. contribution to airquality) and recreation among the local population. However italso increases the value of the land for housing development in theparts of La Primavera with access to the city.

The metropolitan area of Guadalajara has been urbanizedrapidly in recent decades, particularly in the municipalities ofZapopan, Tonala and Tlajomulco. For example, Zapopan's popula-tion has increased by one million between 1970 and 2010 (INEGI,1990, 2000, 2010b). In the same period the Tesistan Valley inZapopan, once one of the most productive agricultural areas inMexico, saw its agricultural area shrink from 28,000 ha to around13,000 ha (Inforural, 2007). Tonala's population grew almosttwenty fold from 1970 to reach 478,789 in 2010 (INEGI, 1990,2000, 2010b) and urbanization buried the best clay mines fromwhich traditional local artisan pottery was made, and for whichthe region is famous (Loera, 2005). Most recently, urbanization hasbeen focused in the southern municipality of Tlajomulco, wherethe population increased at a mean yearly rate of 12.5% from 2000to 2010 (INEGI, 1990, 2000, 2010b). During 1997–2007, around4300 ha of ejido land were sold, mostly to non-ejido buyers in theperi-urban municipalities of the metropolitan area of Guadalajaraand about 4200 ha were privatized (INEGI, 2007).

In regions more distant from the urban area, grazing andcommercial agricultural practices are common (e.g. sugar cane,agave for tequila production and maize). The cumulative effect ofthese land use changes outside the protected core part of thereserve, and the construction of roads around it, is to isolate LaPrimavera and interrupt important natural wildlife corridors toother forested areas, compromising the long-term viability ofpopulations of vulnerable wildlife species (e.g. large wide rangingcarnivores such as the Puma, Puma concolor). The researchpresented here aims to identify the potential of incentive-basedmechanisms to promote the provision of environmental servicesin the region and is part of a broader project funded by the DarwinInitiative, investigating the role that market-based mechanismsmight have in the conservation of wildlife corridors and ruraldevelopment (Balderas Torres et al., 2009).

3. Methods

3.1. Choice modeling

Choice modeling (CM) is a stated preference technique foundedon demand, welfare and consumer theory (Lancaster, 1966) andthe random utility model (RUM) (McFadden, 1974). It was firstdeveloped for marketing and transportation studies and is nowincreasingly used in environmental non-market valuation(Bennett and Blamey, 2001; Hoyos, 2010). It allows the valuationof changes in environmental goods and services, or policiesmodifying their level of provision. In CM, respondents choosebetween hypothetical scenarios that vary in their specific char-acteristics, which are described by different attributes. In this casewe investigated how the payment level, the length of the PEScontract and the offer of other co-benefits increasing local humanand productive capitals affect the preferences of landowners forPES programs.

In CM it is assumed that choices made by landowners max-imize their utility (Bennett and Adamowicz, 2001). The choices areanalyzed in econometric models representing the utility of an

individual. Most econometric models consider an additive utilityfunction where the attributes used in the choice experiment,individual characteristics and alternative associated constants(ASC) describing different scenarios are used as independentvariables (Bennett and Adamowicz, 2001; Hoyos, 2010). TheRUM implies that there is a stochastic component denoted by anerror term. The commonly used multinomial logit model (MNL)assumes that this term follows a Gumbel distribution (Louviere,2001; McFadden, 1974). This assumption imposes the restriction ofindependence of irrelevant alternatives (IIA) that can be diagnosedthrough the Hausman–McFadden test (Hausman and McFadden,1984; Maddala, 1983). Heterogeneity in the sample and tastes ofthe participants may cause violations to the IIA assumption, inwhich case the model may produce biased parameters (Ben-Akivaand Lerman, 1985; Bennett and Adamowicz, 2001). Although theseviolations need to be considered, if the information associatedwith the unobserved components is small, then the effect of anymisspecification in the distribution of errors would be minor(Vojacek and Pecakova, 2010). In these cases the MNL modelwould be preferred to nested logit and probit models, reducingcomputational effort and facilitating the determination of theprobabilities of choosing specific alternatives, thus reducing modelestimation problems (Vojacek and Pecakova, 2010).

CM allows the assessment of trade-offs among the differentattributes of the experiment (Bennett and Adamowicz, 2001). Asone of the attributes in a choice experiment is a monetarycomponent, this allows the estimation of implicit prices, bydividing the parameter of a given attribute by that from themonetary one (Rolfe et al., 2000). In this research we utilize aMNL model with an additive utility function including the attri-butes of the scenarios. For a more detailed description of CMplease refer to Hoyos (2010) or Rolfe et al. (2000). We estimated anMNL model to obtain the implicit valuation of the co-benefitsoffered and the project duration (part worths); and used thebootstrap method of Krinsky and Robb with 7500 iterations tocreate 95% confidence intervals for these values (Krinsky and Robb,1986).

3.2. Survey design and application

The survey posed a hypothetical situation in which a non-governmental organization (NGO) develops projects for forestconservation and rural development. The NGO would receiveresources from private companies and citizens and invite land-owners/ejidos to participate in a PES project. The benefits that theparticipants could receive were a cash payment and a set of co-benefits, over a given period of time. It was proposed that the NGOwould help to improve the health and education services for theparticipants (by paying health insurance fees and hiring highschool teachers for the service of the participants and families).Alternatively the NGO could help participants to get jobs and startsmall business by providing consultancy and financing. Conditionsfor participation in all cases included: (1) a specific forested areashould be conserved or restored and monitored; and (2) measuresshould be taken to prevent and control fires/pests, grazing,hunting and illegal logging. The commitments and responsibilitieswould be specified in a collaboration agreement with the NGO andno penalties were referred to or included.

3.2.1. Choice setsFor the design of the choice sets the research team, including

the authors, two market research technicians and a forester,selected the attributes and levels jointly with local experts (i.e.the managerial team of La Primavera and local researchers andconsultants experienced working in the area). A pilot study was

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–8176

done among key stakeholders and landowners to test the surveyinstrument, and these participants were interviewed to identifytheir interests and needs. Three attributes were chosen: cashpayments, with four different levels ($31, $71, $117 and $165 perhectare per year); project duration (5, 9 or 17 years), and threelevels for the potential co-benefits. The baseline was where no co-benefits were offered in terms of health and education services(level 2) and employment/productive projects (level 3).

An orthogonal main effects design generated 16 choice cards.Twelve choice sets each presenting two project proposals and oneopt out for those deciding not to participate in the program wereprepared in two questionnaires. Questions were designed tomaintain a balance between the different levels of attributes. Forthis, eight cards were included two times. In order to reduce thepresentation of dominant options, the choice sets were preparedaiming to present options generating similar aggregated economicvalues. Because of heterogeneity in property size, we decided notto include an attribute specifying the area of forest to be com-mitted for these practices. Instead, this issue was explored in thefollow-up questions.

3.2.2. ApplicationThe survey was carried out among 161 ejido and non-ejido

landowners located in La Primavera and related wildlife corridorsup to a distance of 80 km from the Reserve. In the study area thereare about 90 ejidos, nine of which were selected at differentdistances from Guadalajara for the study in order to include areaswith different opportunity costs associated with urban develop-ment. Later, in a geographic information system (GIS), we esti-mated the Euclidean distance to the metropolitan area ofGuadalajara. Ejido committees were contacted to present theresearch and invite them to participate. Ejido members werecontacted in their monthly meetings. Non-ejido landowners werecontacted through the regional association of private landownersof Jalisco. The list of representatives in each municipality in thestudy area was consulted to contact the registered landowners.Surveys were applied anonymously, in person, and on anindividual basis.

Before applying the survey, background information and theproposed scenario were described to the participants. The infor-mation presented included, firstly a brief background about thegoods and services produced by forests, and then the problemsdue to deforestation and land use change and initiatives developedthrough public and private PES programs in Mexico and differentcountries. The scenario and conditions for participation were alsodescribed. Participants were then presented with the choice sets.After the choice sets follow-up questions were included to ask ifthe respondent would participate if the project became opera-tional, in which case they were asked how much land they wouldlike to include in the project. They were also asked if they wouldreforest non-forested areas (i.e. agricultural lands and/or grass-lands) in order to integrate them in the project, or if they would bewilling to adopt other agroforestry practices (i.e. living fences,windbreaks). Respondents from ejidos were asked if they wouldagree to include communal forested areas in the project. In orderto get proxies for opportunity costs, we asked about the activitiestheir properties are currently used for, and the expected prices atwhich land is sold in the area where the land is located. Otherquestions captured general socioeconomic and demographic infor-mation. Participants were asked the price of their land in the realestate market as a proxy for land opportunity costs. Surveys werecarried out between March and August 2011, involving landownersin the La Primavera Biosphere Reserve (n¼28) and in the biolo-gical corridors (n¼133); 127 surveys were completed by ejidatar-ios who owned a total area of 2029 ha (median 7 ha per

respondent) and 34 by private landowners who encompassed atotal area of 2686 ha (median of 28 ha per respondent). The totalarea held by the participants was 4715 ha. The total area belongingto the nine ejidos visited was 34,440 ha of which 11,440 ha arecommon land (RAN, 2011). This data reflects the differences in thescale of land ownership in the region; the proportion of ejido toprivate land in the sample is similar to that in the region, howeverthe size of the plots in the ejido regime is far smaller. Assumingthat each ejidatario has only one plot of land or parcela, theselected ejidos would have 5887 members. It is difficult to obtainexact figures on membership because of migration and thedynamic linked with sale and privatization of land as mentionedin Section 2.2. However it is estimated that the number of ejidomembers interviewed represented around 2% of the populationwithin the ejidos selected, which would represent around 0.2% ofall the ejidos in the region. However, cultural, climatological andeconomic conditions are more or less homogenous in the studyarea, apart from distance to the city, which we believe allows theresults to be generalized. Even when participants did not haveforested properties, they were interviewed to investigate theirpreferences and to explore if they would reforest their propertiesin order to participate. When these participants were part of anejido they were asked about the potential enrollment of communalforest areas. Factors affecting the possibilities for reforestation canbe used to analyze the potential to restore wildlife corridors andsequester carbon.

4. Results and discussion

4.1. Sample characteristics

Table 1 presents the basic characteristics of the sample.In our sample 55.9% of the respondents held forested land; the

remaining percentage did not have forested areas but agriculturalparcels and/or grasslands. With respect to current use, 75.8% of therespondents indicated that they used their land for productivecash activities (e.g. sugar cane, agave—tequila, maize); 47.2%indicated that part of the production was also for their ownconsumption in the household and 7.7% had participated in thenational PES program. In our sample 6.3% of the participants werefemales, 9.4% were single and 96.2% were economically active. Theparticipants were, therefore, predominantly relatively old, marriedmales, which is consistent with other studies of Mexican land-owners in rural areas in Mexico and other countries (e.g. in the U.S., Dickinson et al., 2012).

4.2. MNL model and implicit prices

Table 2 presents the MNL model, no violations with the IIAassumption were detected after the Hausman–McFadden test wasapplied; moreover the pseudo R2 value is 0.425 which is consid-ered similar to a coefficient of determination (R2) above 0.6 to0.8 for an equivalent linear regression (Vojacek and Pecakova,2010); thus the estimation of alternative models is not justified.The exercise asks the participants for the willingness to accept in asimple way, since the values are expressed on a per hectare basisand do not require making a specific commitment about whichparticular piece of the parcel might be devoted to the project. Inthis sense, the decision-making process was homogenous and thedifferent alternatives presented were considered to be indepen-dent from each other. An ASC for the status quo alternative wasincluded (opt out), this permits differentiating the effect of theoptions offering only the cash payment from that of the opt-out.Responses of participants who found the questions confusing(3.1%) or those deciding not to participate in the PES in the six

Table 3Implicit prices for the different attributes with 95% Intervals, Krinsky and Robbmethod.

Attribute Implicit prices (95% C.I.)

Duration −$17.04 (−$27.72, −$11.38)No co-benefits (None) −$0.71 (−$56.68, $93.25)Health and education $59.13 ($5.68, $156.17)Employment and productive projects $102.79 ($40.92, $216.75)

Table 1General statistical information of the sample.

Mean S.D. Range

Age (years) 58.1 13.3 20–87Household size (members) 4.3 2.0 1–14Education (Class)a 1.88 1.46 0–5Income (Class)b 2.00 1.02 1–4Income ($/cap-day) 5.10 4.54 1.39–36.43Distance from Guadalajara City (km) 42.4 21.0 0–74.9Land price forest ($) (n¼42) 83,000 331,500 1500–2,200,000Land price non-forest ($) (n¼100) 66,500 269,700 1100–2,600,000Forest area owned (ha) (n¼82) 33.7 75.4 0.2–450Agricultural area owned (ha) (n¼139) 11.6 20.8 0.1–180Grassland (ha) (n¼23) 14.5 15.7 1.0–56Total Area (ha/respondent) 30.6 71.1 0.1–550How much area would you commit if the project was implemented? (ha) 14.5 34.1 0–225

a Education classes: (0) None, (1) Elementary School, (2) Jr. High School, (3) High School, (4) Technical Studies, (5) Undergraduate.b Monthly income classes (1) below $295; (2) $295–$590; (3) $590–$1100; (4) 4$1100.

Table 2MNL model.

Coefficients

ASC status quo (Not to participate) −2.5576***

(0.1307)Payment ($/ha-yr) 0.0069***

(0.0012)Duration (yr) −0.1176***

(0.01155)Co-benefit 1 (None) −0.0049

(0.2452)Co-benefit 2 (Health and education) 0.4080**

(0.1868)Co-benefit 3 (Employment and productive projects) 0.7093***

(0.1843)

Valid (n) 2915pR2 0.425Model fitting information (d.f.) 1010 (6)***

Standard Error in Parenthesis.nnn Significant at 99% level.nn Significant at 95% level.

Table 4Answers to questions exploring the potential for participation.

If the Project becomes operational would you participate: Response (%)

Certainly yes 82.6Probably yes 8.1Probably not 0.6Certainly not (Not answered) 8.7Plans for the future (10 years)Agriculture 71.4Business/Development 8.1Environmentally friendly productive projectsa 22.4Would you support that communal areas join theproject (ejidos)? (n¼127)

97.6

a Ecotourism, vermicomposting, reforestation.

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–81 77

choice sets (protests) (6.2%) were not included following standardmethods (e.g. Glenk and Colombo, 2011; Scarpa et al., 2009).

The pseudo-R2 is within the recommended values and thecoefficients obtained are highly significant and with the expectedsigns, i.e. higher payments and shorter project periods are pre-ferred. The projects offering co-benefits are more highly valuedthan the option where only the cash payment is offered (thecoefficient for the option with no co-benefit is close to zero andstatistically not significant).

Considering the coefficients obtained in Table 2 as the proxy forprobability of participation, the odds of joining the PES increase byabout 0.7% per dollar offered per hectare per year. Respondentspreferred shorter projects to the certainty of predictable paymentsover larger periods of time; this is in agreement with other casesreported in the literature (e.g. Markowski-Lindsay et al., 2011).Offering health and education co-benefits increased the chances ofparticipation by 40.8%, on the other hand offering aid for employ-ment and productive projects increased the chances by 70.9%. Theimplicit prices for the different project attributes are presented inTable 3 by considering the magnitude of the coefficients for eachattribute and of the payment.

By considering the inverse of the coefficient of the payment, itcan be seen that if a project were to rely only on cash payments,the yearly payment per hectare should be $144.92 for the odds onparticipating to increase to100%. The additional payment required

to increase the chances of participation in a project, includingsupport for employment and productive projects, would be of$42.13/ha-year, since the valuation of the co-benefits offeredunder this option already accounts for $102.79. For a scheme thatprovides health and education benefits the average requiredpayment would be $85.79.

4.3. Opportunity costs and potential participation

Table 4 shows the potential for participation among respon-dents across the whole sample if the project proposed becameoperational. There is high potential for participation among land-owners, although 9.3% expressed overall negative opinions onparticipation. The potential for participation was explored inrelation to the distance to Guadalajara (Fig. 2). Responses suggestthat when landowners are located close to the urban area, thehigher opportunity costs of land would have a negative effect onprospects for participation. These landowners have better accessto services and economic opportunities offered in the city, whichwould reduce the relative value of any in-kind co-benefits offered.

Fig. 2. Potential participation as function of distance to the urban area limit. Thedashed lines represent the trend expected in the peri-urban area between LaPrimavera and the city of Guadalajara where forested areas have been cleared toopen space for other land uses.

Fig. 3. Land price and distance to the urban area (note logarithmic scale of the landprice axis). The dashed lines represent the trend expected in the peri-urban areabetween La Primavera and the city of Guadalajara, where forested areas have beencleared to open space for other land uses.

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–8178

The two curves in Fig. 2 are complementary and fit an inverserelationship well, indicating that landowners very close to the cityare less likely to participate, i.e. the probability of participationincreases with distance. Although no interviews were made inareas closer than 8 km to the city it can be seen that between theprotected area and the western urban limit there is a buffer ofaround 5 km where there are no remaining forests (Fig. 1). Theserevealed decisions over land use support the specification formspresented in this figure. The regressions imply that at a distance of5 km or less from the city there would be no interest in participat-ing in PES; however the dashed lines are included to delimit thevalid distance range for the data in the sample. The forest cover inLa Primavera has been preserved thanks to the status of NationalProtected Area and Biosphere Reserve. This status prevented aproposed plan to create a large housing development within thereserve during the 1970s (Ciudad Primavera). Nevertheless privatelandowners and ejidos were not compensated in any form whenLa Primavera was declared a protected area. Pressures for devel-opment still persist; and there are various cases of illegal con-struction of country houses inside the Reserve that are underlitigation (del Castillo, 2011). The data reported by respondents onland price was also classified by distance in order to assess landopportunity costs (Fig. 3).

The data shows a power relationship between distance andland price indicating that the pressure of urban development willbe far more intense in areas closer to the city. This is in agreementwith the implications of Fig. 2. The opportunity cost of land in thevicinity of the city would be extremely high in comparison withthe incentives that could be offered by voluntary programs such asPES. The pro-development context prevailing in this area, in termsof access to services, markets and formulation of urban develop-ment plans, is correlated with lower expected participation inincentive based programs (e.g. Markowski-Lindsay et al., 2011). Asin Fig. 2, there are no data points for distances below 5 km thatcould confirm the curve specification in areas closer to the city.Again, the dashed gray lines are included to describe the generaltrend that could be expected within the most proximate area tothe city outside the distance range of the data collected. Althoughin practice it is unreasonable to expect that price for land will tendto infinity, it does increase considerably more in those areas closerto the city when urbanization has been authorized. Moreover, thefunction represents a price trend similar to other cases where the

value of scarce resources has been estimated, (e.g. Costanza et al.,1997, assumed that the price of an environmental resource willtend to infinity when its supply approaches zero). However theequation in Fig. 3 performs well in estimating the opportunitycosts associated with land price within the valid distance range.

Although the prices of new housing developments may not becomparable to land prices, since they include the cost of construc-tion, from Fig. 3 it can be seen that even if the new houses weresold at the construction cost, the profit made when land bought atlow prices is sold at the peri-urban market price is potentiallyenormous. Speculation in the real-estate market might contributeto the disordered peri-urban growth experienced in the last fewdecades. Effective controls over land use change, particularly closeto urban areas, are an essential precondition for the provision ofenvironmental services; otherwise the landowners' expectationsof urban development will prevent participation in schemesaiming to value them. This is particularly critical in hotspot areasfor biodiversity such as La Primavera and its unprotected wildlifecorridors, given its proximity to Guadalajara. It is worth mention-ing that land prices for development are affected by governmentdecisions over where houses can or cannot be built, thus they arenot pure market prices as usually considered for efficiency. Never-theless some land use regulations can be changed to respond tothe economic interests of landowners and allow development. Inareas designated for conservation, high land prices can be anindication of possible conflicts and pressures for land use conver-sion that may threat biodiversity.

4.4. Implications of offering higher payments in PES

Landowner's participation in PES could be increased if highercash payments are offered in areas with higher opportunity costs.However if higher payments are offered it can be the case that thenet present value of the incentives paid can overcome land prices(Sullivan et al., 2005). This can compromise long term efficiency ofthe program if high enrollment rates are required, since purchaseof land for conservation may be the least cost option in areas withlow opportunity costs (Fig. 4).

In Fig. 4 the horizontal lines represent the land prices perhectare for the minimum land price reported, the lower 15%percentile and the median along with the net present value ofpayments in PES for different project durations as a function ofyearly payment levels. The net present values are estimated

Fig. 4. Net present value of PES for different project durations and discount rates asa function of yearly payment and comparison with land prices.

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–81 79

without discounting, and using a discount rate of 7%. Discountrates from 4 to 6% have been used to estimate the net presentvalue of incentives for forest banking (Sullivan et al., 2005); herewe use a slightly higher figure to account for a lower economicdevelopment. The minimum price of land documented is equal-ized by PES projects paying between $55 to 250/ha-year depend-ing on contract length and discounting; the payments made bysome PES programs in Mexico are within this range (e.g. PROBOS-QUE). Likewise the net present value of PES would match the valueof the 15% percentile of land prices for payments in the range of$300 to $1250/ha-year.

Buying land for conservation may be a more cost efficient optionin the long term in some cases, but will be highly controversial. Inrural areas, especially in those located away from cities, land is oneof the most important productive, social and cultural assets. Thecomparatively lower land prices will be influenced by the localpoverty context (Martínez-Alier, 2004). If the payments set are afunction of opportunity costs, they will systematically offer lowerincentives in poor areas. Hence the search for lower payments whenaiming for cost efficiency should be carefully assessed; especiallywhen considering the benefits and savings that project intermedi-aries and buyers/users of the services may be obtaining (i.e. incarbon offsetting markets). The relative balance between cash andin-kind payments in incentive-based mechanisms should be alsocarefully balanced in such a way that the interests of the commu-nities/landowners providing the environmental services and thoseof the financing bodies/users of those services are fulfilled; in anycase the importance of each type of benefit could be negotiated aspart of the PES agreement itself. How such win-win negotiations canbe achieved in cases where the power relations may be veryasymmetric is an important area for further research. Resultspresented here imply that although full opportunity costs are notcovered, the value of pro-development interventions may increasethe probabilities of enrollment. This would be more notable in ruralareas away from cities where providing enhanced access to servicesassociated with social, human and productive capital and employ-ment would be more valued.

Although the general purchase of land for conservation can becontested on political, social and economic grounds, it must beborne in mind that when opportunity costs of land increase (dueto urban expansion, or creation of infrastructure), landownersmight sell the land. This is demonstrated by the process ofurbanization in Mexico after the constitutional reform and thecase of Guadalajara, as mentioned in Section 2.2. Under these

circumstances, strict land use regulation is required otherwisevoluntary participation in incentive-based mechanisms will not besufficient to ensure provision. Purchase of land for provision ofenvironmental services can offer a long-term solution in specific,limited, fragile or hotspot areas in exceptional cases e.g. wildlifecorridors.

4.5. Potential enrollment into PES in the study region

Apart from the effect of Guadalajara's urban opportunity costs,there is a positive potential for participation in PES in forestedareas in the region. This is particularly noticeable for ejido com-munal areas, as 97.6% of the responses were positive about theirinclusion into PES. In the ejidos were the surveys were appliedthese areas represent one third of the ejido land, and there are upto around 100,000 ha in the region (RAN, 2011). When participantswere asked if they would reforest areas not currently forested (i.e.private land and individual ejido parcels), 26.8% answered posi-tively indicating they would afforest grasslands that are used forgrazing. However, this could have been affected by a drought in2011 which reduced livestock profitability (Toledo, 2012). Agricul-tural areas would not be afforested: in the region, prices for sugarcane were around $48/ton and for maize $400/ton and yields were120 t/ha and 10 t/ha respectively. However the potential foragroforestry is high since 96.4% of those owning agricultural areaswere positive about planting trees as windbreaks. This indicatesthat agroforestry practices may be attractive for landowners sincethey do not require full land conversion and may thus represent aviable option to enhance certain environmental services (e.g.carbon sequestration) (Balderas Torres et al., 2010). Afforestationof grasslands and establishment of windbreaks can be fundedthrough alternative sources of income in carbon markets forforestry offsets.

Other than the external pressures of urban opportunity costs,factors that need to be considered for potential applicability ofthese results to other sites are related to the type of ecosystemsand management practices developed by the forest-holders. For-ests in dry environments similar to those in the La Primaveraregion offer relatively few options for productive uses that targetcash-markets, the most common being seasonal agriculture andgrazing. In more humid areas, ejidos and landowners may bealready engaged in productive forestry practices which may not beeligible for participation in PES. Moreover, at the community levelcertain institutional conditions and local social capital are requiredfor successful implementation (e.g. Bray et al., 2006). Neverthelessresults from the surveys are mirrored by experiences observed inthe national PES program and on-going projects in the voluntarycarbon markets. Most of the land committed to PES by ejidos inMexico corresponds to communal-use areas (Muñoz Piña et al.,2008). On the other hand agroforestry practices, such as wind-breaks and shade coffee, are commonly practiced in the projectsdeveloped in the voluntary carbon markets (De Jong et al., 1995;Balderas Torres et al., 2010). In some areas they may represent upto 90% of committed land, as demonstrated by the Scolel Té projectin Chiapas (Quechulpa-Montalvo, personal communication). Thisindicates that, in general, preference for inclusion of common-useareas and agroforestry practices under PES can be extrapolated toother sites once particular conditions (e.g. opportunity costs fromurban development) are taken into account. Naturally this willdepend on whether such forest management practices are con-sidered eligible under the given PES scheme.

Currently there is a PES-like project under design in the LaPrimavera area based on the results of the project funded by theDarwin Initiative. The project includes the conservation of about1100 ha, and reforestation of 100 ha more in one of the wildlife

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–8180

corridors. It also includes pro-development interventions (i.e. adevelopment community center, tree nursery, computer centerand high school facilitation). The budget required for ten years ofoperation of this project is around $125/ha-yr. This is in line withthe higher end PES programs in Mexico. Moreover it is importantto note that, although the associated pro-development activitiesrepresent an important share of the costs (close to 50%), the socialbenefits expected from them can offset and exceed the overallcosts of the project.

This research describes the general characteristics of an incen-tive based mechanism designed to finance conservation practicesin the region. In the study area (Fig. 1), the development of a PESscheme targeting existing forested areas would require $16.3million per year at the current payment level of the national PESprogram, ($22.2/ha-yr) or close to $90 million at the rate expectedfor the pilot project. Pro-development interventions could benefitover 250,000 persons living in around 400 rural settlements in theregion (i.e. settlements of 100 to 2500 inhabitants). The totalpopulation in the area, excluding the metropolitan area of Gua-dalajara is around 720,000 (INEGI, 2010b). Maximum potentialimplementation for windbreaks through agroforestry projectswould be around 112,000 km (considering the perimeter ofaverage parcels of 4 ha). If one quarter of the grasslands werereforested this would represent around 40,000 ha of new forests.For the purposes of biodiversity conservation, particular attentionshould be given to the areas linking La Primavera to other forestedareas as they are critical for the movement of wildlife; althoughthese may represent only around a thousand hectares. The longterm viability for these projects depends on the capacity to createthe financing mechanisms to value the environmental servicesprovided by the ecosystems and improved management practices.If these mechanisms adequately target demand, then there will bemore chances of generating sustainable financial flows (Pagiola etal., 2002). The potential participation of landowners in incentivebased schemes, as presented here, needs to be complemented byan assessment of the potential environmental services produced inthe territory (e.g. hydrological, landscape, carbon sequestration,biodiversity), a study of the demand and valuation of theseservices among different users (i.e. local in Guadalajara and non-local in the carbon markets), and the creation of an institutionalframework to link the producers and users (e.g. a public PESprogram or projects focused on the carbon markets).

5. Conclusions

Rural areas tend to maintain greater levels of natural capitalwhile urban areas concentrate assets and expertise around finan-cial, productive and institutional human and social capitals. PESand market mechanisms address this dichotomy, focusing on thevaluation of natural capital to link the users of the environmentalservices in developed areas with the providers in rural areas.These interventions are devised to deal with the environmentalexternalities commonly considered by economists to representmarket failures. However poverty can also be seen as a failure togenerate and maintain adequate levels of human, productive,social and financial capitals. An integrated strategy to valorizeenvironmental services could include mechanisms not only toensure the flow of natural services from rural to developed areas,but also to transfer resources specifically linked to human, social,productive and financial capitals into rural areas appropriate to thelocal contexts, while ensuring the provision of environmentalservices. Results of this study show that landowners prefer PESschemes including co-benefits that increase local levels of humanand productive capital. Including the co-benefits may reduce theimportance and dependence on cash payments and would

increase the chances for participation. However we consider thatthe mix of cash and in-kind compensation should be negotiated bythe financing body and local providers as part of the agreementsought in a market based mechanism. The potential participationdiminishes in the presence of high opportunity costs from urbanareas. In these areas benefits and co-benefits offered by PESprojects might not suffice to prevent forest conversion into otherland uses. Thus other policy options, such as land use restrictions,monitoring and sanctions should be considered to capture thevalue of environmental services in these areas.

Acknowledgments

The authors thank M.G., I.G., M.A.O.M, A.A.M., J.O., F.Q.M. andpersonnel from ITESO and La Primavera Executive Office for theircontribution during fieldwork, to R.O.E. and J.P.G. for the assistancein producing the maps and to two anonymous reviewers for theircomments that helped to improve the previous version of themanuscript. The research was funded by the Darwin Initiative(Project: 17027 to J.C.L. and A.B.T.). A.B.T. acknowledges scholar-ships granted by CONACYT and SEP.

References

Balderas Torres, A., Lovett, J.C., Skutsch, M., 2009. Assessing the feasibility to linkurban and rural areas through local markets for forest's carbon services and thepotential for local development: a methodological proposal. In: Proceedings ofthe XIII World Forestry Congress. Forest in Development a Vital Balance.Buenos Aires, Argentina.

Balderas Torres, A., Marchant, R., Lovett, J.C., Smart, J.C.R., Tipper, R., 2010. Analysisof the carbon sequestration costs of afforestation and reforestation agroforestrypractices and the use of cost curves to evaluate their potential for implementa-tion of climate change mitigation. Ecological Economics 69, 469–477.

Balvanera, P., Cotler, H., 2007. Acercamientos al estudio de los servicios ecosisté-micos, vols. 84–84. Gaceta Ecológica. SEMARNAT, Mexico, 8–15.

Barnes, G., 2009. The evolution and resilience of community-based land tenure inMexico. Land Use Policy 26, 393–400.

Ben-Akiva, M.E., Lerman, S.R., 1985. Discrete Choice Analysis: Theory and Applica-tion to Travel Demand. MIT Press, Cambridge, USA.

Bennett, J., Blamey, R., 2001. Introduction. In: Bennett, J., Blamey, R. (Eds.), TheChoice Modelling Approach to Environmental Valuation. Edward Elgar, Chel-tenham, UK, pp. 1–10.

Bennett, J., Adamowicz, V., 2001. Some fundamentals of environmental choicemodelling. In: Bennett, J., Blamey, R. (Eds.), The Choice Modelling Approach toEnvironmental Valuation. Edward Elgar, Cheltenham, UK, pp. 37–72.

Bray, D.B., Antinori, C., Tores-Rojo, J.M., 2006. The Mexican model of communityforest management: the role of agrarian policy, forest policy and entrepreneur-ial organization. Forest Policy and Economics 8, 470–484.

Brouwer, R., Tesfaye, A., Pauw, P., 2011. Meta-analysis of institutional-economicfactors explaining the environmental performance of payments for watershedservices. Environmental Conservation 38 (4), 380–392.

Canadell, J.G., Raupach, M.R., 2008. Managing forests for climate change mitigation.Science 320, 1456.

Carney, D., 1998. Approaches to Sustainable Livelihoods for the Rural Poor. ODIPoverty Briefings 2, London.

Church, A., Ravenscroft, N., 2008. Landowner responses to financial incentiveschemes for recreational access to woodlands in South East England. LandUse Policy 25, 1–16.

CONANP, 2000. Natural Protected Area Management Plan for La Primavera.Comisión Nacional de Áreas Naturales Protegidas, SEMARNAT, Mexico.

Corbera, E., Kosoy, N., Martínez Tuna, M., 2007. Equity implications of marketingecosystem services in protected areas and rural communities: case studies fromMeso-America. Global Environmental Change 17, 365–380.

Cornelius, W., Myhre, D., 1998. The Transformation of Rural Mexico: Reforming theEjido Sector. Center for US-Mexican Studies. University of California, San Diego.

Costanza, R., et al., 1997. The value of the world's ecosystems and natural capital.Nature 387, 253–260.

Daw, T., Brown, K., Rosendo, S., Pomeroy, R., 2011. Applying the ecosystem servicesconcept to poverty alleviation: the need to disaggregate human well-being.Environmental Conservation 38 (4), 370–379.

de Groot, R.S., 1992. Functions of Nature: Evaluation of Nature in EnvironmentalPlanning, Management and Decision Making. Wolters-Noordhoff, Groningen.

de Groot, R.S., Wilson, M.A., Boumans, R.M.J., 2002. A typology for the classification,description and valuation of ecosystem functions, goods and services. Ecologi-cal Economics 41, 393–408.

A. Balderas Torres et al. / Ecosystem Services 6 (2013) 72–81 81

De Jong, B.H.J., Montoya-Gomez, G, Nelson, K., Soto-Pinto, L., Taylor, J., Tipper, R.,1995. Community forest management and carbon sequestration: a feasibilitystudy from Chiapas, Mexico. Interciencia 20, 409–416.

del Castillo, A., 2011. La Primavera, una defensa imposible. Local NewspaperMilenio. 10 October, 2011. Available online: ⟨http://impreso.milenio.com/node/9041077⟩ (accessed 8.04.12).

Dickinson, B.J., Stevens, T.H., Markowski-Lindsay, M., Kittredge, D.B., 2012. Esti-mated participation in U.S. carbon sequestration programs: a study of NIPFlandowners in Massachusetts. Journal of Forest Economics 18, 36–46.

Dupraz, P., Vanslembrouck, I., Bonnieux, F., van Huylenbroeck, G., 2003. Farmers'Participation in European Agri-Environmental Policies. Paper presented at 10th

Congress of the European Association of Agricultural Economists, Zaragoza, Spain.Engel, S., Pagiola, S., Wunder, S., 2008. Designing payments for environmental

services in theory and practice: an overview of the issues. Ecological Economics65, 663–674.

Estado de México, 2011. Convocatoria para el pago por servicios ambientaleshidrológicos 2011. Gaceta del Gobierno. Periódico Oficial del Gobierno delEstado Libre y Soberano de México, No. 27. Toluca de Lerdo, México. 8 deFebrero de 2011.

Glenk, K., Colombo, S., 2011. Designing policies to mitigate the agriculturalcontribution to climate change: an assessment of soil based carbon sequestra-tion and ancillary effects. Climate Change 105, 43–66.

Greiner, R., Herr, A., Stoeckl, N., 2003. Policies Aimed at Overcoming Impedimentsto the Implementation of On-farm Conservation Activities. Paper presented at4th BioEcon Workshop on the Economics of Biodiversity Conservation, Venice.

Grieg-Gran, M., Porras, I., Wunder, S., 2005. How can market mechanisms forforests environmental services help the poor? Preliminary lessons from LatinAmerica. World Development 33 (9), 1511–1527.

Hausman, J.A., McFadden, D., 1984. Specification tests for the multinomial logitmodel. Econometrica 52, 1219–1240.

Horne, P., 2004. Forest owners' acceptance of incentive-based policy instruments inforest biodiversity conservation: a choice experiment approach. Silva Fennica40 (1), 169–178.

Hoyos, D., 2010. The state of the art of environmental valuation with discrete choiceexperiments. Ecological Economics 69, 1595–1603.

INEGI, 1990. Censo de población y vivienda 1990. Instituto Nacional de Estadística yGeografía. Aguascalientes, México.

INEGI, 2000. Censo de población y vivienda 2000. Instituto Nacional de Estadísticay Geografía. Aguascalientes, México.

INEGI, 2005. Delimitación de las zonas metropolitanas de México. InstitutoNacional de Estadística y Geografía. Aguascalientes, México.

INEGI, 2007. Censo Agropecuario 2007, IX Censo Ejidal. Instituto Nacional deEstadística y Geografía. Aguascalientes, México.

INEGI, 2010a. Conjunto nacional de uso de suelo y vegetación a escala 1:250,000Serie IV. Instituto Nacional de Estadística y Geografía. Aguascalientes, México.

INEGI, 2010b. Censo de población y vivienda 2010. Instituto Nacional de Estadísticay Geografía. Aguascalientes, México.

Inforural, 2007. Produce Valle de Tesistán más maíz en menor superficie. Availableonline: ⟨http://www.inforural.com.mx/spip.php?article14519⟩ (accessed on80.4.12).

Kilgore, M.A., Snyder, S.A., Schertz, J., Taff, S.J., 2008. What does it take to get familyforest owners to enroll in a forest stewardship-type program? Forest PolicyEconomics 10, 507–514.

Kinzig, A.P., Perrings, C., Chapin III, F.S., Polasky, S., Smith, V.K., Tilman, D., Turner II,B.L., 2011. Paying for ecosystem services—promise and peril. Science 334,603–604.

Krinsky, I., Robb, A., 1986. Approximating the statistical properties of elasticities.Review of Economics and Statistics 68, 715–719.

Kroeger, T., Casey, F., 2007. An assessment of market-based approaches to providingecosystem services on agricultural lands. Ecological Economics 64, 321–332.

Lambert, D.M., Sullivan, P., Claassen, R., Foreman, L., 2007. Profiles of US farm householdsadopting conservation-compatible practices. Land Use Policy 24, 72–88.

Lancaster, K., 1966. A new approach to consumer theory. Journal of PoliticalEconomy 74, 132–157.

Landell-Mills, N., Porras., I.T., 2002. Silver Bullet or Fool's Gold: A Global Review ofMarkets for Forest Environmental Services and their Impact in the Poor.International Institute for Environment and Development, London, England.

Loera, M.E., 2005. Artesanos en peligro de extincion. Gaceta Universitaria 394,18–19.

Loftus, T.T., Kraft, S.E., 2003. Enrolling conservation buffers in the CRP. Land UsePolicy 20 (1), 73–84.

López-Hoffman, L., Varady, R.G., Flessa, K.W., Balvanera, P., 2010. Ecosystem servicesacross borders: a framework for transboundary conservation policy. Frontiersin Ecology and the Environment 8, 84–91.

Louviere, J., 2001. Choice experiments: an overview of concepts and issues. In:Bennett, J., Blamey, R. (Eds.), The Choice Modelling Approach to EnvironmentalValuation. Edward Elgar, Cheltenham, UK, pp. 13–36.

Lynch, L., Hardie, I., Parker, D., 2002. Analyzing Agricultural Landowners'Willingness to Install Streamside Buffers. Working Paper 02-01. Departmentof Agricultural and Resource Economics, The University of Maryland, CollegePark.

Ma, Z., Butler, B.J., Kittredge, D.B, Catanzaro, P., 2012. Factors associated withlandowner involvement in forest conservation programs in the U.S.: implica-tions for policy design and outreach. Land Use Policy 29, 53–61.

MacMillan, D.C, Phillip, S., 2010. The role of economic incentives in resolvingconservation conflicts: the case of wild deer management and habitat con-servation. Human Ecology 38 (4), 485–494.

Maddala, G.S., 1983. Limited-Dependent and Qualitative Variables in Econometrics.Cambridge University Press, USA.

Markowski-Lindsay, M., Stevens, T., Kittredge, D.B., Butler, B.J., Catanzaro, P.,Dickinson, B.J., 2011. Barriers to Massachusetts forest landowner participationin carbon markets. Ecological Economics 71, 180–190.

Martínez-Alier, J., 2004. The Environmentalism of the Poor: A Study of EcologicalConflicts and Valuation. Oxford University Press, New Delhi.

McFadden, D., 1974. Conditional logit analysis of qualitative choice behavior. In:Zarembka, P. (Ed.), Frontiers in Econometrics. Academic Press, New York, USA,pp. 105–142.

Muñoz Piña, C., Guevara, A., Torres, J.M., Braña, J., 2008. Paying for the hydrologicalservices of Mexico's forests: analysis, negotiations and results. EcologicalEconomics 65, 725–736.

Muradian, R., Corbera, E., Pascual, U., Kosoy, N., May, P.H., 2010. Reconciling theoryand practice: an alternative conceptual framework for understanding paymentsfor environmental services. Ecological Economics 69, 1202–1208.

Newton, P., Nichols, E.S., Endo, W., Peres, C.A., 2012. Consequences of actor levellivelihood heterogeneity for additionality in a tropical forest payment forenvironmental services programme with an undifferentiated reward structure.Global Environmental Change 22, 127–136.

Pagiola, S., Bishop, J., Landell-Mills, N., 2002. Making market-base mechanismswork for forest and people. In: Pagiola, S., Bishop, J., Landell-Mills, N. (Eds.),Selling Forest Environmental Services: Market-based Mechanisms for Conser-vation and Development. Earthscan, London. (Chapter15).

Pagiola, S., Platais, G., 2007. Payments for Environmental Services: from Theory toPractice. World Bank, Washington, USA.

Petrzelka, P., Malin, S., Gentry, B., 2012. Absentee landowners and conservationprograms: mind the gap. Land Use Policy 29, 220–223.

Putten, v.I.E., Jennings, S.M., Louviere, J.J., Burgess, L.B., 2011. Tasmanian landownerpreference for conservation incentive programs: a latent class approach.Journal of Environmental Management (92), 2647–2656.

RAN, 2011. Polígonos de Areas Ejidales. Registro Agrario Nacional, México.Reyes-González, J.A., Gómez-Barrón, J.P., Osaland-Muis, R., R., Zavala Gómez del

Campo, 2012. Potencial de servicios ambientales en la propiedad social enMéxico. México-Instituto Interamericano de Cooperación para la Agricultural.Secretaría de la Reforma Agraria, México.

Rico Garcia-Amado, L., Ruiz Perez, M., Reyes Escutia, F., Barrasa Garcia, S., ContrerasMejia, E., 2011. Efficiency of payments for environmental services: equity andadditionality in a case study from a Biosphere Reserve in Chiapas, Mexico.Ecological Economics 70, 2361–2368.

Rolfe, J., Bennett, J., Louviere, J., 2000. Choice modelling and its potential applicationto tropical rainforest preservation. Ecological Economics 35 (2), 289–302.

Scarpa, R., Gilbride, T.J., Campbell, T., Hensher, D.A., 2009. Modelling attribute non-attendance in choice experiments for rural landscape valuation. EuropeanReview of Agricultural Economics 36 (2), 151–174.

Scullion, J., Thomas, C.W., Vogt, K.A., Pérez-Maqueo, O., Logsdon, M.G., 2011.Evaluating the environmental impact of payments for environmental servicesin Coatepec (Mexico) using remote sensing and on-site interviews. Environ-ment Conservation 38 (4), 426–434.

Stevens, T.H., White, S., Kittredge, D.B., Dennis, D., 2002. Factors affecting NIPFlandowner participation in management programs: a Massachusetts casestudy. Journal of Forest Economics 8 (3), 169–184.

Sullivan, J., Amacher, G.S., Chapman, S., 2005. Forest banking and forest landownersforgoing management rights for guaranteed financial returns. Forest Policy andEconomics 7, 381–392.

Toledo, A., 2012. Provoca sequía ventas de pánico ganadero. Local NewspaperMilenio. 9 January, 2012. Available online: ⟨http://jalisco.milenio.com/cdb/doc/impreso/9091958⟩ (accessed 8.04.12).

UNESCO, 2011. Man and the Biosphere Programme. United Nations Educational,Scientific and Cultural Organization. ⟨http://www.unesco.org/mabdb/br/brdir/directory/contact.asp?code=MEX⟩ (accessed 29.09.11).

Vojacek, O., Pecakova, I., 2010. Comparison of Discrete Choice Models for EconomicEnvironmental Research. Prague Economic Papers, vol. 1, pp. 35–53.

Viana, V.M., 2008. Bolsa Floresta: um instrumento innovador para a promoção da saúdeem comunidades tradicionais na Amazônia. Estudos Avançados 22 (64), 143–153.

Wunder, S., 2005. Payments for Environmental Services: Some Nuts and Bolts.Occasional Paper No. 42 CIFOR, Bogor, Indonesia.